Science Forum

Science Forum

  1. Joined
    06 Mar '12
    Moves
    625
    17 Jun '19 10:45
    https://phys.org/news/2016-10-ultrasound-imaging-gaining-precision.html

    This link doesn't give any meaningful details of how that software works and, as a (modest) computer expert, I would really like to know how it works. All it say on it is;
    "They developed 'smart' echographic image-reconstruction algorithms that interpret the echoes picked up by the probe and fill in the missing data.".
    I found this sun-link that explains what I want to know;
    https://ieeexplore.ieee.org/document/7582552
    But, unfortunately, its language is far to technical even for me so I had to given up trying.

    But at least I (and we) now know that significant improvements CAN be made to imaging just by making some improvements in the image-processing software alone, which I personally think is nice to know.
    Not sure but perhaps some part of my current AI research will eventually help with that? -we will just have to wait and see for that one.
  2. Joined
    06 Mar '12
    Moves
    625
    17 Jun '19 11:45
    My above missedit; "sun-link" should be "sub-link"
  3. Standard memberDeepThought
    Losing the Thread
    Cosmopolis
    Joined
    27 Oct '04
    Moves
    80670
    17 Jun '19 13:10
    @humy said
    https://phys.org/news/2016-10-ultrasound-imaging-gaining-precision.html

    This link doesn't give any meaningful details of how that software works and, as a (modest) computer expert, I would really like to know how it works. All it say on it is;
    "They developed 'smart' echographic image-reconstruction algorithms that interpret the echoes picked up by the probe and fill in the ...[text shortened]... current AI research will eventually help with that? -we will just have to wait and see for that one.
    I read the abstract of your second link. It looks like they do Fourier analysis on the output and have some way of removing artifacts. I need to "sign in" to read the whole thing and will make a decision about that later as my last major job was helping to write and improve a VSIPL function library - that Vector Signal and Image Processing. The functions were things like FFTs (Fast Fourier Transforms), Digital Filters, and Linear Algebra functions (solvers and so forth). So thanks for putting me on to this.
  4. Joined
    06 Mar '12
    Moves
    625
    17 Jun '19 14:156 edits
    @deepthought said
    I read the abstract of your second link. It looks like they do Fourier analysis on the output and have some way of removing artifacts. I need to "sign in" to read the whole thing and will make a decision about that later as my last major job was helping to write and improve a VSIPL function library - that Vector Signal and Image Processing. The functions were things li ...[text shortened]... Filters, and Linear Algebra functions (solvers and so forth). So thanks for putting me on to this.
    I am afraid I think all that is over my head because I had to look up both "Vector Signal and Image Processing" and "Fast Fourier Transforms" as these things wasn't anything I did at university with my many computer courses. Just me taking a quick peek at the wiki formula definition of "Fast Fourier Transforms" (see https://en.wikipedia.org/wiki/Fast_Fourier_transform ), which involves complex numbers, instantly makes me give up trying to understand it (at least for now).
    I think I better only spend time trying to understand only what I need to understand to do my current unique and highly specialized AI research, which will take me a very long time and is certainly hard enough to understand!
    You not only know more about physics than I do but you seem to know more about computer science than I do despite computer science being my main area of specialty. Anyone who says or believes you aren't very well formally qualified in science is truly an idiot (guess who I have in mind here).
Back to Top